Ascl2 knockdown results in tumor growth arrest by miRNA-302b-related inhibition of colon cancer progenitor cells

Abstract

Background: Achaete scute-like 2 (Ascl2), a basic helix-loop-helix (bHLH) transcription factor, controls the fate of intestinal stem cells. However, the role of Ascl2 in colon cancer progenitor cells remains unknown. The cell line HT-29 (47.5-95% of CD133(+) population) and LS174T (0.45% of CD133(+) population) were chosen for functional evaluation of Ascl2 in colon cancer progenitor cells after gene knockdown by RNA interference.

Methodology/principal findings: Immunohistochemistry demonstrated that Ascl2 was significantly increased in colorectal adenocarcinomas. Downregulation of Ascl2 using RNA interference in cultured colonic adenocarcinoma HT-29 and LS174T cells reduced cellular proliferation, colony-forming ability, invasion and migration in vitro, and resulted in the growth arrest of tumor xenografts in vivo. The Ascl2 protein level in CD133(+) HT-29 cells was significantly higher than in CD133(-) HT-29 cells. Ascl2 blockade via shRNA interference in HT-29 cells (shRNA-Ascl2/HT-29 cells) resulted in 26.2% of cells staining CD133(+) compared with 54.7% in control shRNA-Ctr/HT-29 cells. The levels of 'stemness' associated genes, such as CD133, Sox2, Oct4, Lgr5, Bmi1, and C-myc, were significantly decreased in shRNA-Ascl2/HT-29 and shRNA-Ascl2/LS174T cells in vitro as well as in the corresponding tumor xenograft (CD133 was not performed in shRNA-Ascl2/LS174T cells). The shRNA-Ascl2/HT-29 cells had inhibited abilities to form tumorspheres compared with control. The microRNA (miRNAs) microarrays, identified 26 up-regulated miRNAs and 58 down-regulated miRNAs in shRNA-Ascl2/HT-29 cells. Expression levels of let-7b, miRNA-124, miRNA-125b, miRNA-17, miRNA-20a and miRNA-302b, involved in the regulation of 'stemness', were quantified with qPCR, which confirmed their identities. Restoration of miRNA-302b, via its mimic, led to the restoration of shRNA-Ascl2/HT-29 'stemness' characteristics, including tumorsphere formation and 'stemness' associated genes levels, and the recovery of cellular behaviors, including colony-forming ability, invasion and migration in vitro.

Conclusions/significance: Ascl2 may be a potential target for the inhibition of colon cancer progenitor cells, and functions through a miR-302b-related mechanism.

Figure 1. Expression of Ascl2 in normal human colon mucosa, colon cancer and colon cancer cell lines and its selective interference in HT-29 and LS174T cells.

Ascl2 expression is specifically localized at the nucleus of crypt base cells of normal human colon mucosa (arrowhead) (A), B demonstrates that Ascl2 is specifically expressed at the nucleus of human colon cancer cells in human colon cancer tissues (Original magnification of top panel of A and B: ×200; Original magnification of low panel of A and B: ×400). The immunofluorescent staining indicates Ascl2 located mainly in the nucleus of HT-29 cells and the cytoplasm of LS174T cells (C) (Original magnification: ×200). Western blot analysis shows Ascl2 is present in HT-29 and LS174T cells, but absent in MHCC-97L cells (D). Ascl2 interference in HT-29 and LS174T cells results in the significant reduction of both Ascl2 mRNA analyzed by real-time PCR and protein levels analyzed by western blot analysis relative to control (β-actin) (**: p<0.01) (E and F).

Figure 2. Ascl2 blockade in HT-29 and LS174T cells results in the inhibition of colony formation, proliferation, invasion and migration in vitro.

shRNA-Ascl2/HT-29 and shRNA-Ascl2/LS174T cells have fewer colonies (*: p<0.05) (A), lower growth rates (*: p<0.05) (B), less invaded cells through the Matrigel-coated membrane (**: p<0.01) (C) and less cells migrating across the scraped edge (*: p<0.05) (D), when compared with their controls (Original magnification: ×200).

Figure 3. Ascl2 interference in HT-29 and LS174T cells leads to tumor growth arrest in vivo.

All mice (6/6, respectively) develop tumors 20 days later after 1×10⁶ shRNA-Ctr/HT-29 cells (right side and marked as arrow) and shRNA-Ascl2/HT-29 cells (left side and marked as arrowhead) are inoculated into nude mice (A), LS174T cells were not shown. The tumor volume (B) and mass weight (C) in the group of shRNA-Ascl2/HT-29 and shRNA-Ascl2/LS174T cells is significantly lower than the group of shRNA-Ctr/HT-29 and shRNA-Ctr/LS174T cells (*: p<0.05). The mRNA (D) and protein (E) levels of Ascl2 in the tumor tissues develop from shRNA-Ascl2/HT-29 and shRNA-Ascl2/LS174T cells are lower than in the tumor tissues developed from shRNA-Ctr/HT-29 and shRNA-Ctr/LS174T cells (**: p<0.01). Ascl2 immunostaining in the nucleus of the cancerous cells in tumor xenografts from shRNA-Ctr/HT-29 and shRNA-Ctr/LS174T cells is stronger than in the nucleus of the cancerous cells in tumor xenografts from shRNA-Ascl2/HT-29 and shRNA-Ascl2/LS174T cells (F).

Figure 4. Sorting of HT-29 cells based on CD133 marker and Ascl2 expression in CD133⁺ and CD133⁻ HT-29 cells.

CD133 is expressed in 58.1% of HT-29 cells. 98.6% of cells are confirmed to be CD133 positive by post-sorting selection in the top 10.8% of CD133⁺ HT-29 cells, 98.1% of cells are confirmed to be CD133 negative by post-sorting selection in the dimly 7.6% of CD133⁻ HT-29 cells (A). Ascl2 protein level relative to control (β-actin) in CD133⁺ HT-29 cells is significantly higher than that in CD133⁻ HT-29 cells (*: p<0.05) (B). There is an obvious nuclear staining of Ascl2 in CD133⁺ HT-29 cells, but Ascl2 is almost negative in CD133⁻ HT-29 cells (C) (Original magnification: ×200).

Figure 5. The percentage of CD133⁺ HT-29 cells and the expression level of ‘stemness’ associated genes are reduced due to Ascl2 knockdown in vitro.

54.7% of shRNA-Ctr/HT-29 cells are positive for CD133 expression compared with 26.2% of HT-29 cells are positive for CD133 expression in shRNA-Ascl2/HT-29 cells (*: p<0.05) (A and B). The mRNA levels of ‘stemness’ associated genes, like CD133 (not performed in LS174T cells and its transfectants), Lgr5, Oct4, Bmi1, Sox2, and C-myc analyzed by real-time PCR in the shRNA-Ascl2/HT-29 and shRNA-Ascl2/LS174T cells are lower than in the shRNA-Ctr/HT-29, HT-29, shRNA-Ctr/LS174T and LS174T cells, respectively (*: p<0.05; **: p<0.01) (C). The protein levels of ‘stemness’ associated genes, like CD133 (not performed in LS174T cells and its transfectants), Oct4, Bmi1, Sox2, and C-myc, analyzed by western blot analysis in shRNA-Ascl2/HT-29 and shRNA-Ascl2/LS174T cells are significantly lower than that in the shRNA-Ctr/HT-29, HT-29, shRNA-Ctr/LS174T and LS174T cells, respectively (D).

Figure 6. The expression level of ‘stemness’ associated genes are reduced due to Ascl2 knockdown in vivo.

The mRNA levels of CD133 (not performed in the tumor tissues developed from shRNA-Ascl2/LS174T and shRNA-Ctr/LS174T), Lgr5, Oct4, Bmi1, Sox2, and C-myc, analyzed by real-time PCR in the tumor tissues developed from shRNA-Ascl2/HT-29 and shRNA-Ascl2/LS174T cells are lower than in the tumor tissues developed from shRNA-Ctr/HT-29 and shRNA-Ctr/LS174T cells (*: p<0.05; **: p<0.01) (A). The protein levels of CD133 (not performed in the tumor tissues developed from shRNA-Ascl2/LS174T and shRNA-Ctr/LS174T), Oct4, Bmi1, Sox2, and C-myc, analyzed by western blot analysis in the tumor tissues developed from shRNA-Ascl2/HT-29 and shRNA-Ascl2/LS174T cells are significantly lower than in the tumor tissues developed from shRNA-Ctr/HT-29 and shRNA-Ctr/LS174T cells (B).

Figure 7. Knockdown of Ascl2 leads to the inhibition of the tumorsphere formation.

1000 cells of shRNA-Ascl2/HT-29, shRNA-Ctr/HT-29 and non-transfected HT-29 cells are plated for tumorsphere formation as described in Materials and methods and quantified at 100× and 400× magnifications (A). The number of tumorspheres from shRNA-Ascl2/HT-29 cells is significantly less than that from shRNA-Ctr/HT-29 and non-transfected HT-29 cells (*: p<0.05) (B). The number of cells per tumorsphere from shRNA-Ascl2/HT-29 cells is significantly less than that from shRNA-Ctr/HT-29 cells and non-transfected HT-29 cells (*: p<0.05) (C).

Figure 8. Transfection of miR-302b mimic in shRNA-Ascl2/HT-29 cells restores their ‘stemness’ characteristics and recovers their cellular behaviors in vitro.

The let-7b, miRNA-124 miRNA-125b are significantly up-regulated, the miRNA-302b, miRNA-20a and miRNA-17 are significantly down-regulated, in shRNA-Ascl2/HT-29 cells compared with shRNA-Ctr/HT-29 cells (*: p<0.05;**: p<0.01) (A). The native shRNA-Ascl2/HT-29 cells, shRNA-Ascl2/HT-29 cells transfected with NC mimic and shRNA-Ascl2/HT-29 cells transfected with miR-302b mimic were analyzed for tumorsphere formation and were quantified at 100× (left panel of B) and 400× (right panel of B) magnifications. The number of tumorspheres and cells per tumorsphere from shRNA-Ascl2/HT-29 cells transfected with miR-302b mimic were significantly higher than those from shRNA-Ascl2/HT-29 and shRNA-Ascl2/HT-29 cells transfected with NC mimic (*: p<0.05) (C). Western blot analysis of Ascl2, C-myc, CD133, Bmi1, Sox2 and Oct4 in the cell lysates indicate that Ascl2, Sox2 and Oct4 protein levels are induced due to miR-302b mimic transfection in shRNA-Ascl2/HT-29 cells compared with shRNA-Ascl2/HT-29 and shRNA-Ascl2/HT-29 cells transfected with NC mimic (D). The real time PCR experiments for quantification of Ascl2, Oct4 and Sox2 mRNAs demonstrate a significant increase due to miR-302b mimic transfection of shRNA-Ascl2/HT-29 cells (*: p<0.05) (E). The colony-forming ability, the numbers of invaded cells and migrated cells of shRNA-Ascl2/HT-29 cells transfected with miR-302b mimic were significantly increased comparing with control (*: p<0.05) (F).